Back to Ecosystem Pulse
OperationsReference
ignite25-PREL13-observe-manage-and-scale-agentic-ai-apps-with-microsoft-foundry
by microsoft
Hands-on notebooks for observing and scaling agentic apps with Azure Foundry
Jupyter Notebook
Updated Mar 26, 2026
Share:
Overview
Demonstrates how to observe, manage, and scale agentic AI applications using Azure AI Foundry. Walks through hands-on notebooks that instrument agent interactions and surface observability signals, and apply scaling and governance patterns on Azure services. Includes examples connecting Foundry models to Azure OpenAI and Azure Cognitive Search for retrieval-augmented agent workflows.
Key Benefits
As agents become system components, visibility into their interactions and reliability is essential for trust and governance. These notebooks make observability and operational patterns concrete—showing how to log agent-to-agent exchanges, monitor failure modes, and apply continuous evaluation in an Azure environment. That practical focus helps teams turn abstract ideas like agent track record and continuous evaluation into reproducible operational practices.
When to Use
Teams prototyping production-ready agent systems on Azure who need concrete observability, scaling, and governance examples.
Applications
- Instrument agent-to-agent interactions to collect logs and trace signals for debugging
- Implement continuous quality and safety checks for agent outputs in an Azure pipeline
- Scale agent workloads using Azure Foundry patterns and integrate RAG with Azure Cognitive Search
- Prototype governance and observability workflows before production rollout
Topics
agent-evaluationaiopsazure-ai-foundryazure-ai-foundry-modelsazure-ai-searchazure-openaidistillation-modelobservabilityquality-evaluationsafety-evaluation+1 more
Similar Tools
agent-playgroundautogen
Keywords
multi-agent trustagent track recordproduction agent monitoringazure-ai-foundry